Visuomotor Origins of Covert Spatial Attention Tirin Moore, Katherine M Armstrong, Mazyar Fallah  Neuron  Volume 40, Issue 4, Pages 671-683 (November 2003) DOI: 10.1016/S0896-6273(03)00716-5

Figure 1 Presaccadic “Reactivation” of a V4 Neuron The location of area V4 on the surface of the prelunate gyrus is shown in the side view of the macaque cerebral cortex. The response of the V4 neuron was recorded during a visually guided delayed saccade task. While the monkey fixated a central spot (FP), a visual target was presented inside the cell's receptive field (RF). After a delay, the fixation point was extinguished and the monkey made a saccade to the target. The eye position during each trial is shown by a representative horizontal position trace. The response of a single V4 neuron is illustrated in the raster below the event plots. The occurrence of a single action potential is denoted by a dot in the raster during each of 11 trials arranged vertically and aligned to the onset of the saccade. Note that not only is this neuron activated initially when the target first appears but also just before the saccade is made to the RF stimulus. (Adapted from Fischer and Boch, 1985.) Neuron 2003 40, 671-683DOI: (10.1016/S0896-6273(03)00716-5)

Figure 2 Activity of a V4 Neuron during Visually Guided Saccades, and the Degree of Visual Guidance of Saccades to Targets of Different Orientation (A) The responses of a single V4 neuron during delayed saccadic eye movements to an oriented bar stimulus presented within a V4 cell's RF. Each plot shows the mean instantaneous firing rate of the cell when either the preferred (top) or nonpreferred (bottom) orientation was used as the target. The cell's activity during the first half of trials is aligned to the onset of the RF stimulus (left) and during the second half is aligned to the saccade (right). The cartoon above illustrates the two phases of the trial. In the first half, the point of fixation (“+”) was on the fixation spot, and in the second half, the monkey made a saccade to the RF stimulus (arrow). When the preferred stimulus was presented, the neuron responded not only at the time of onset but also immediately before the monkey made a saccade to the stimulus. When the bar stimulus was presented at the nonpreferred orientation, the neuron failed to respond robustly at either time. (B) Saccades made to oriented bar stimuli during the visually guided, delayed saccade task. Shown are individual saccade vectors (black lines) obtained from a single block of trials (13 trials/orientation). Note that the saccade end points fall along the axis of bar orientation. Neuron 2003 40, 671-683DOI: (10.1016/S0896-6273(03)00716-5)

Figure 3 Overt and Covert Visual Attention (A) Example of the overt selection of visual features contained within a portrait of a young woman as revealed by the scanning eye movements of a human subject. The figure on the right shows the subject's eye position traces superimposed on the portrait displayed by itself on the left. (Adapted from Yarbus, 1967.) (B) An example of covert attention in macaque monkeys. The female monkey on the left is looking straight ahead and past the large threatening male on the right. However, it is apparent that the smaller monkey is fully attending to her neighbor. Neuron 2003 40, 671-683DOI: (10.1016/S0896-6273(03)00716-5)

Figure 4 Overt and Covert Spatial Attention Effects in Area V4 (A) In studies of overt attention (left), a monkey is trained to make saccades to stable visual stimuli after a delay, and the effects on the presaccadic responses are compared between the different conditions. The event plot at the top indicates the period during which the activity of the cell is compared in the attended (saccade to RF) and unattended (saccade away) conditions (tick marks at time of saccade). The bottom plots summarize the basic results obtained, namely that overall visual responses are enhanced prior to saccades that target the RF stimulus and that the enhancement is greatest for the most effective stimuli (Fischer and Boch, 1981a, 1981b; Moore et al., 1998; T. Moore, A.S. Tolias, and P.H. Schiller, 1998, Soc. Neurosci., abstract). In studies of covert attention (right), monkeys are trained to attend to a peripheral stimulus while maintaining central fixation (spotlight). The attended stimulus can be placed within the RF of the V4 neuron under study, or at some other location, and the visual responses compared. The event plot at the top indicates the period over which the activity of the cell is compared in the attended and unattended conditions (tick marks on stable eye position). The bottom plots summarize the basic results obtained: overall visual responses are enhanced when the attention is directed to the RF stimulus (e.g., Motter, 1993) and that this enhancement is greatest for the most effective stimuli (McAdams and Maunsell, 1999). (B) At the time overt attention (a saccade) is directed to a location outside of but near a cell's RF, the RF of some cells shifts toward the target location (Tolias et al., 2001). When covert attention is directed to a location outside of but near a cell's RF, the RF of some cells shifts toward the attended location (gray ring) (Connor et al., 1997). Neuron 2003 40, 671-683DOI: (10.1016/S0896-6273(03)00716-5)

Figure 5 Microstimulation of the FEF and Its Effects on Covert Spatial Attention The location of the FEF is shown in gray in the side view of the macaque brain. (A) Electrical stimulation of FEF sites can evoke saccades that shift a monkey's gaze to a fixed location with respect to the center of gaze (MF, or “movement field”). (B) The probability that a saccade will be evoked varies as a logistic function of current amplitude. In the example depicted, a saccade is evoked 50% of the time when the threshold current of 25 μA is applied to the FEF site. (C) Monkeys were trained on an attention task in which they had to detect the transient dimming of a visual stimulus (target) while ignoring a flashing distracter that appeared sequentially at random locations of the display. The target was placed at the location to which the monkey's gaze would be shifted with suprathreshold stimulation of the FEF site (part A). The attention task was performed with and without subthreshold stimulation of the FEF site on randomly interleaved trials. On stimulation trials, microstimulation occurred immediately prior to the dimming of the target stimulus. (D) Microstimulation increased the monkey's sensitivity to the target change, relative to nonstimulation trials. The distribution of relative sensitivity values (microstimulation/control, expressed logarithmically) is shifted above 0 (mean at arrow), indicating increased sensitivity during microstimulation. This effect was only observed when the target was positioned within the MF. (Adapted from Moore and Fallah, 2001, 2003.) Neuron 2003 40, 671-683DOI: (10.1016/S0896-6273(03)00716-5)

Figure 6 Effect of FEF Microstimulation on the Visual Response of V4 Neurons (A) Microstimulation of sites within the FEF was carried out while simultaneously recording the responses of single V4 neurons to visual stimuli in monkeys performing a fixation task. The FEF microelectrode was positioned so as to align the evoked saccade vector with the RF position of the V4 cell under study (bottom cartoon). (B) (Top) Example of the effect of FEF microstimulation on the response of a single V4 neuron to an oriented bar presented to the cell's RF (cartoon above) when the saccade vector represented at the FEF site (arrow) overlapped with the V4 RF. Mean response during control trials is shown in black, and the mean response of trials on which a 50 ms microstimulation train (FEF stim) was applied to the FEF site is shown in red. (Bottom) Same as in top, but histograms show responses during trials on which a visual stimulus is only presented outside of the RF. The response of the cell was elevated immediately following the stimulation train but only when the cell was being driven by a RF stimulus (i.e., top versus bottom histogram). (C) The stimuation-driven enhancement of the cell's response depended critically on the effectiveness of the visual stimulus. When there was no RF stimulus, there was a near zero change in the cell's response, compared to control trials. When there was a nonoptimally oriented vertical bar in the RF, there was an intermediate enhancement. The greatest enhancement was observed when a horizontally oriented bar stimulus was presented inside the RF. (Adapted from Moore and Armstrong, 2003.) Neuron 2003 40, 671-683DOI: (10.1016/S0896-6273(03)00716-5)